New environmentally friendly oxidative scission of oleic acid into azelaic acid and pelargonic acid

Article abstract of DOI:10.1007/s11746-012-2134-7

Oleic acid (OA) is a renewable monounsaturated fatty acid obtained from high oleic sunflower oil. This work was focused on the oxidative scission of OA, which yields a mono-acid (pelargonic acid, PA) and a di-acid (azelaic acid, AA) through an emulsifying system. The conventional method for producing AA and PA consists of the ozonolysis of oleic acid, a process which presents numerous drawbacks. Therefore, we proposed to study a new alternative process using a green oxidant and a solvent-free system. OA was oxidized in a batch reactor with a biphasic organic-aqueous system consisting of hydrogen peroxide (H ₂O₂, 30 %) as an oxidant and a peroxo-tungsten complex Q₃{PO₄[WO(O₂)₂]₄} as a phase-transfer catalyst/co-oxidant. Several phase-transfer catalysts were prepared in situ from tungstophosphoric acid, H₂O₂ and different quaternary ammonium salts (Q⁺, Cl^-). The catalyst [C₅H₅N(n-C₁₆H₃₃)] ₃{PO₄[WO(O₂)₂]₄} was found to give the best results and was chosen for the optimization of the other parameters of the process. This optimization led to a complete conversion of OA into AA and PA with high yields (>80 %) using the system OA/H ₂O₂/[C₅H₅N(n-C₁₆H ₃₃)]₃{PO₄[WO(O₂)₂] ₄} (1/5/0.02 molar ratio) at 85 C for 5 h. In addition, a new treatment was developed in order to recover the catalyst.

Full text of DOI:10.1007/s11746-012-2134-7

J Am Oil Chem Soc (2013) 90:133–140
DOI 10.1007/s11746-012-2134-7
ORIGINAL PAPER
New Environmentally Friendly Oxidative Scission of Oleic Acid
into Azelaic Acid and Pelargonic Acid
•
•
•
Ana¨ıs Godard Pascale De Caro Sophie Thiebaud-Roux
•
Emeline Vedrenne Zephirin Mouloungui
´
Received: 15 March 2012 / Revised: 24 May 2012 / Accepted: 9 August 2012 / Published online: 26 August 2012
Ó AOCS 2012
Abstract Oleic acid (OA) is a renewable monounsatu-
rated fatty acid obtained from high oleic sunflower oil. This
work was focused on the oxidative scission of OA, which
yields a mono-acid (pelargonic acid, PA) and a di-acid
(azelaic acid, AA) through an emulsifying system. The
conventional method for producing AA and PA consists of
the ozonolysis of oleic acid, a process which presents
numerous drawbacks. Therefore, we proposed to study a
new alternative process using a green oxidant and a sol-
vent-free system. OA was oxidized in a batch reactor with
a biphasic organic-aqueous system consisting of hydrogen
peroxide (H2O2, 30 %) as an oxidant and a peroxo–tung-
sten complex Q3{PO4[WO(O2)2]4} as a phase-transfer
catalyst/co-oxidant. Several phase-transfer catalysts were
prepared in situ from tungstophosphoric acid, H2O2 and
different quaternary ammonium salts (Q?, Cl–). The cata-
lyst [C5H5N(n-C16H33)]3{PO4[WO(O2)2]4} was found to
give the best results and was chosen for the optimization of
the other parameters of the process. This optimization led
to a complete conversion of OA into AA and PA with high
yields ([80 %) using the system OA/H2O2/[C5H5N(n-
C16H33)]3{PO4[WO(O2)2]4} (1/5/0.02 molar ratio) at 85 °C
for 5 h. In addition, a new treatment was developed in
order to recover the catalyst.
Keywords Hydrogen peroxide ꢀ Oleic acid ꢀ Oxidative
cleavage ꢀ Peroxo–tungsten complex ꢀ Phase-transfer
catalysis ꢀ Solvent-free reaction ꢀ Tungstophosphoric acid
Abbreviations
AA
EO
Azelaic acid
Ethyl oleate
FAME Fatty acids methyl esters
LA
OA
PA
RA
RT
Linoleic acid
Oleic acid
Pelargonic acid
Ricinoleic acid
Room temperature
TBME tert-Butyl methyl ether
TMSH Trimethylsulfonium hydroxide
TPA
Tungstophosphoric acid
Introduction
Monounsaturated fatty acids are renewable starting mate-
rials that are particularly attractive for the chemical
industry because of their low price and worldwide avail-
ability. Among them, oleic acid or cis-9-octadecenoic acid
(OA) is a C18:1 monounsaturated fatty acid, which can be
obtained from high oleic sunflower oil. This renewable
resource is commonly grown in the south of France and in
other Mediterranean countries. OA thus represents an
interesting building block for the chemical industry [1].
The most common C=C transformations of OA are
epoxidation, hydroxylation and oxidative cleavage. The
oxidative scission is particularly attractive as it provides a
C9 mono-acid (pelargonic acid or nonanoic acid, PA) and a
A. Godard ꢀ P. De Caro (&) ꢀ S. Thiebaud-Roux ꢀ
E. Vedrenne ꢀ Z. Mouloungui
´
Universite de Toulouse, INPT-ENSIACET, LCA (Laboratoire
de Chimie Agro-Industrielle), ENSIACET, 4 Allee Emile
´
Monso, 31030 Toulouse, France
e-mail: Pascale.DeCaro@ensiacet.fr
A. Godard ꢀ P. De Caro ꢀ S. Thiebaud-Roux ꢀ E. Vedrenne ꢀ
Z. Mouloungui
INRA, UMR 1010 CAI, 31030 Toulouse, France
123

134
J Am Oil Chem Soc (2013) 90:133–140
C9 di-acid (azelaic acid or nonanedioic acid, AA). These
saturated acids exhibit short and odd hydrocarbon chains
that are present in very low amounts in natural resources.
Their rarity is all the more problematic since these acids
are very interesting starting materials for the preparation of
numerous bio-based products [1, 2]. AA is, for example,
converted into different esters for the preparation of poly-
mers (Nylon 6:9), plasticizers, adhesives, solvents, biode-
gradable lubricants, corrosion inhibitors, and anti-acneic
agent for cosmetics [1, 3, 4]. PA is an intermediate in the
production of lubricants, plasticizers, perfumes, herbicides,
fungicides, resins [1, 5].
Experimental Procedures
Materials and Reagents
High oleic sunflower oil was purchased from Iterg (Pessac,
France). Aqueous hydrogen peroxide (aqueous solution,
30 %), tungstophosphoric acid (H3PW12O40ꢀnH2O, TPA,
reagent grade), linoleic acid (LA, w/w 62.0 %), formic acid
(97–98 %), AliquatÒ 336, cetylpyridinium chloride
monohydrate, tetrabutylammonium chloride (C97 %), tet-
rabutylammonium bromide (99 %), potassium iodide
(99 %), ethyl acetate (technical grade) and tert-butyl
methyl ether (TBME, reagent grade) were purchased from
Sigma Aldrich Co. Ammonium heptamolybdate tetrahy-
drate and sulfuric acid solution (4 N) were provided from
VWR. Candida cylindracea lipase, trimethylsulfonium
hydroxide (TMSH) and sodium thiosulfate solution (0,
1 N) were respectively purchased from Lyven, Macherey–
Nagel and Fluka.
Several oxidant systems have been developed to cleave
the double bond of OA or methyl oleate using chromic acid,
nitric acid, potassium permanganate, osmium tetroxide,
peracids, periodates, or hypochlorites [6]. However, most of
these conventional oxidants are not eco–friendly and the
desired compounds are obtained in low yields. Another
oxidant, ozone, was chosen for the large-scale production of
AA and PA from OA. This reaction is very selective and
gives high yields, but it also presents numerous economical
and environmental drawbacks (energy and oxidant con-
sumption, toxicity, explosion hazard) [7].
Hydrolysis of High Oleic Sunflower Oil
High oleic sunflower oil (22.5 kg) was hydrolyzed using
Candida cylindracea lipase in distilled water (20.1 kg)
under stirring at 40 °C for 5 h [12]. The fatty acid com-
position of the hydrolysate was determined by GLC (w/w):
87.6 % oleic acid, 4.7 % linoleic acid, 3.5 % palmitic acid,
3.1 % stearic acid, 0.2 % capric acid and 0.9 % of others
fatty acids. It was used as the source of oleic acid for all
experiments.
In this context, the aim of this work was to propose a new
alternative process using a greener and safer oxidant sys-
tem, such as hydrogen peroxide, as it produces only water as
a waste product [8]. We thus decided to develop new con-
ditions combining its use with a phase-transfer catalyst,
which would enhance the contact between the organic
substrate and the hydrophilic oxidant. Several phase-trans-
fer catalysts were prepared in situ from tungstophosphoric
acid, hydrogen peroxide and different quaternary ammo-
nium salts (Q?, Cl-). The resulting catalysts are peroxo–
tungsten complexes Q3{PO4[WO(O2)2]4} that are thought
to act as co-oxidants [1]. The peroxo anion is the active
ionic species able to transfer oxygen to the organic phase
thanks to its association with a lipophilic cation.
Synthesis of Peroxo–Tungsten Catalytic Complexes
TPA was pre-treated with hydrogen peroxide to form the
peroxo–tungsten catalytic complex. Two methods were
tested.
To our knowledge, a triphasic system OA/H2O2/
Q3{PO4[WO(O2)2]4}, without any organic solvent, had
already been used to epoxidize the double bond of OA [6,
9] but less to perform its cleavage [1, 10, 11, 17]. This
system was a strategic choice in order to develop a green
process, as no organic solvent is used for the reaction itself
and only benign wastes are obtained (if the workup solvent
is recycled). Furthermore, others novelties of this work are
the use of a couple of ammonium quaternary salts ([n-
Bu4N]?, X-; [(n-C8H17)4N]?, X-) which have never been
tested for the cleavage of OA with the selected system, and
the development of a more efficient workup in order to
improve the yields and recycle the catalyst.
Method a: Catalytic complexes were synthesized prior
to the reaction according to a known procedure [10].
Method b: An in-situ protocol was developed for our
study. A solution ammonium quaternary salt (3.6 mmol) in
water (2 mL) was added dropwise to OA in a three-necked
flask. A pre-stirred solution (30 min at room temperature,
RT) of a 30 % aqueous solution of H2O2 (34 mL,
325.0 mmol) and H3PW12O40ꢀnH2O (TPA, 4 g, 1.2 mmol)
in water (5 mL) was then added dropwise.
The following peroxo–tungsten catalytic complexes A,
B, C, D were formed in situ:
A: [n-Bu4N]3{PO4[WO(O2)2]4}
B: [C5H5N(n-C16H33)]3{PO4[WO(O2)2]4}
The influence of various parameters was investigated,
and the reaction conditions were optimized to obtain a
complete conversion and high yields.
C: [MeN(n-C8H17)3]3{PO4[WO(O2)2]4}
D: [(n-C8H17)4N]3{PO4[WO(O2)2]4}
123

J Am Oil Chem Soc (2013) 90:133–140
Oxidative Cleavage of Oleic Acid
135
pelargonic acid (YPA) and of some intermediate products,
such as aldehydes (nonanal and 9-oxo-nonanoic acid) and
epoxide (9,10-epoxyoctadecanoic acid).
Typically, a 250-mL, three-necked, round-bottom flask was
charged with 21 g (65.0 mmol) of OA and the catalyst A,
B, C or D (1.2 mmol). This reactor was equipped with a
condenser, a mechanical stirrer and a thermocouple. The
reaction medium was heated at 60 °C and a 30 % aqueous
solution of H2O2 (325.0 mmol) was added dropwise over
5 min. The reaction mixture was then heated at 85 °C
while stirring (400 rpm). After 5 h, this mixture was
allowed to cool down to room temperature. The pH was
adjusted to 1 with an aqueous hydrochloric acid solution
(4 N). A green ethyl acetate solvent (which can be bio-
sourced) (100 mL) was then added and the reaction mix-
ture was cooled to the desired temperature (T \ 4 °C or
RT). For catalysts A and B, the white precipitate was fil-
tered and washed with ethyl acetate (50 mL). The organic
and aqueous layers were separated in a 500-mL separating
funnel. The aqueous phase was extracted twice with ethyl
acetate (100 mL). The combined organic layers were dried
with anhydrous sodium sulfate (Na2SO4) and evaporated
under reduced pressure. Fatty acids were obtained and
analyzed without a further purification of the crude
material.
The structures of the resulting compounds were identi-
fied by FT-IR spectroscopy (PerkinElmer), ammonia
1
chemical ionization mass spectrometry (MS), H and 13C
NMR (Bruker).
Titration of Hydrogen Peroxide
The hydrogen peroxide amount was titrated with an iodo-
metric method. The aqueous solution containing hydrogen
peroxide (1 mL) was diluted with 100 mL of distilled
water, 30 mL of sulfuric acid (4 N), 35 mL of potassium
iodide (100 g L-1), 1 mL ammonium molybdate (5 g L-1).
The diiodine (I2) formed was titrated with a sodium thio-
sulfate solution (0.1 N).
Results and Discussion
Effects of the Type of Catalyst/Co-oxidant
In order to gain more insight into the oxidative cleavage of
OA using the Q3{PO4[WO(O2)2]4}/H2O2 system, the
influence of the nature of the catalyst was first studied.
A blank experiment, without any catalyst, gave a
mediocre conversion of OA and low AA and PA yields
(Table 1). GLC analysis confirmed the presence of known
intermediate products [10] such as the aldehydes or the diol
and the epoxide formed on the double bond of OA:OA was
partially oxidized but not to the desired products, showing
that a catalyst is needed for this oxidative reaction.
A substrate/oxidant/catalyst molar ratio of 1/5/0.02 was
then chosen for the other experiments. The peroxo–tungsten
complex Q3{PO4[WO(O2)2]4} was formed using a quater-
nary ammonium salt (Q?, X-), TPA and H2O2, following
the protocol described in the experimental procedures
(method b). Four quaternary ammonium cations Q? were
tested in order to form various peroxo–tungsten complexes
with hydrogen peroxide: n-Bu4N? (a), C5H5N(n-C16H33)?
(b), MeN(n-C8H17)?3 (c), (n-C8H17)4N? (d).
Epoxidation of Oleic Acid
The epoxide (9,10-epoxyoctadecanoic acid) was prepared
from OA (71.0 mmol) in freshly made performic acid,
according to a known protocol [13]. The reaction pro-
ceeded at RT for 3 h (400 rpm). The 9,10-epoxyocta-
decanoic acid was obtained as a colorless and powdery
solid in 72 % yield (purity of 65 %, 51.0 mmol).
Analytical Techniques
The collected samples were analyzed and quantified by gas
chromatography on a Varian instrument coupled to a flame
ionization detector (FID). The silica capillary column (CP
for FAME fused silica WCOT, 50 m, 0.25 mm, 0.25 lm)
was purchased from Varian (USA). The carrier gas was
helium (Air Liquide, France) at a pressure of 1.03 bar at
the head of the column. Injector and detector temperatures
were 250 °C. The temperature in the oven was held at
100 °C for 5 min, then ramped to 180 °C at 5 °C min-1 for
10 min and finally increased to 250 °C at 10 °C min-1 for
5 min (held 43 min).
The addition of catalyst Q3{PO4[WO(O2)2]4} led to the
complete conversion of OA, for all the tested catalysts.
This demonstrated that the catalyst is required to transfer
the oxygen of the hydrogen peroxide to the substrate. In
fact, the peroxo anion is the active ionic species able to
transfer oxygen to the organic phase thanks to its associ-
ation with a lipophilic cation. Then, in the aqueous phase,
the catalyst (thus under the oxo form) was reoxidized by
H2O2 into the peroxo form [1].
All samples (10 mg mL-1 dissolved in TBME, 50 lL)
were converted to FAME with TMSH (50 lL). Penta-
decanoic acid was added as the internal standard (2 mg
mL-1).
This analytical procedure allowed us to calculate the
conversion of OA (vOA), the yields of azelaic acid (YAA),
Furthermore,amongthoseexperiments, catalystBgavethe
best yields. This can be explained by the formation of an
123

136
J Am Oil Chem Soc (2013) 90:133–140
Table 1 Activity of different catalytic peroxo complexes, obtained from Q?, X- and TPA, on the oxidative cleavage of OA
Catalyst
Catalyst preparation method
X-
vOA (%)
YAA (%)
YPA (%)
–
b
b
a
–
38.1
100.0
100.0
100.0
100.0
100.0
100.0
100.0
100.0
1.8
77.6
52.0
71.8
81.5
70.2
68.0
75.7
73.2
2.2
80.9
57.1
76.2
86.1
75.1
73.3
80.7
76.5
A
A
A
B
B
B
C
D
Cl-
Cl-
Br-
Cl-
Cl-
Cl-
Cl-
Cl-
b
b
a
a*
b
b
* H2O2 (420 mmol, 6.5 equiv.)
observable emulsion during the reaction. The total interface
area is more important with this catalyst, as it allows the
interfacial tension to be lowered. Moreover, the emulsion is
stabilized by the presence of fatty acids as described as sur-
factants [5]. The medium thusemulsified, favorsthe contact at
theinterfacebetweenthereactantsintheorganicphaseandthe
oxidant in the aqueous phase. The quaternary ammonium
cation of catalyst B, which is very accessible because of the
planar geometry of the pyridinium scaffold, efficiently
transfers the peroxo anion to the organic phase [14].
Compared to methods a and a*, yields of AA and PA
were considerably increased with the in-situ formation of
the peroxo–tungsten complex (method b). Several experi-
ments were run to understand those results. First, the
amount of H2O2 needed for the preparation of the catalyst
using method a was determined by the H2O2 titration left in
the aqueous filtrate: 120 mmol for 1.2 mmol of prepared
catalyst. With method b, a total amount of 650 mmol
(325 ? 325 mmol) of H2O2 is added to the reaction mix-
ture. The oxidant left at the end of the reaction was deter-
mined by titration and was 110 mmol. As 120 mmol should
be needed to form the catalyst (as in method a), it means that
420 mmol (6.5 equiv.) was consumed during the intrinsic
oxidation (oxidant consumption: 80 %). Method a* which
uses the same amount of H2O2 that method b did not
improve yields of AA and PA. Therefore, the efficiency of
the in situ method comes from the method of preparation of
the catalyst.
The influence of the counter-ions was then investigated
by the substitution of the chlorine atom by bromine for
catalyst A. AA and PA yields were lower (4–6 %) using
the bromide-based catalyst. Indeed, the higher reducing
power of the bromide salts compared to the chloride can
generate the partial decomposition of the peroxo–tungsten
complex according to the following reaction: W2O5 +
2Brꢁ + 2Hþ ! Br2 + 2WO2 + H2O [10].
Effects of the Reaction Temperature
Effects of the Catalyst Preparation
The influence of the reaction temperature was then studied
for the three peroxo catalysts A, B and C (Table 2).
For catalysts A and B, the reaction was found to be
optimal between 85 and 90 °C, whereas, for the catalyst C,
it was between 90 and 95 °C. At a lower temperature
(65 °C), there was still some starting OA, even after 5 h of
reaction and a significant amount of intermediate products
was found: epoxide (1.9 %), diol (\2 %) and aldehydes
(14 %) along with a minor amount of non-quantified
compounds. Raising the temperature above 90–95 °C
Peroxo–tungsten complexes can be either synthesized prior
to the reaction [10] (method a) or in situ with a solvent for
the epoxidation [15]. In this study, the peroxo–tungsten
complex was formed in situ but without any organic sol-
vent, which, to our knowledge, has never been described
for such a cleavage reaction (method b). Different proto-
cols (methods a, a*, b) were compared using catalysts A
and B (Table 1). In all cases, a large excess of H2O2 was
used as it has already been described that a catalytic or
thermal decomposition of H2O2 occurs [1].
123

J Am Oil Chem Soc (2013) 90:133–140
137
Table 2 Effect of the reaction temperature on the oxidative cleavage
of OA with A, B and C
Catalyst
Temperature (°C)
vOA (%)
YAA (%)
YPA (%)
A
A
A
B
B
B
B
B
C
C
C
C
C
75
85
100.0
100.0
99.5
59.2
77.6
65.8
40.8
66.8
81.5
77.6
60.9
69.2
75.7
78.5
71.5
72.8
66.5
80.9
73.3
44.2
71.1
86.1
80.2
60.5
74.8
80.7
82.0
76.1
77.2
95
65
96.6
75
100.0
100.0
100.0
99.3
85
95
105
75
Fig. 1 Effects of the agitation speed on the oxidative cleavage of OA
with the catalyst B. YAA (filled rectangles), YPA (filled triangles)
100.0
100.0
99.8
85
95
the yields of AA, PA and intermediate products decreased.
For example, there was no aldehyde left using 4 mol% of
catalyst. Besides, the presence of oligomers was detected
with MS analyzes (m/z = 332, 689 and 861).
105
115
100.0
100.0
Reaction conditions: OA (65.0 mmol), H2O2 (325.0 mmol, 30 %),
catalyst (1.2 mmol), 5 h, 400 rpm
Effects of the Mechanical Stirring Speed
generated a decrease in the desired products yields. This
behavior is likely due to either the degradation of hydrogen
peroxide at high temperature or the oligomerization of the
expected acids. MS analysis confirmed the second
hypothesis, as compounds exhibiting high molecular
weights were detected in the crude reaction mixture
(m/z = 332, 376, 587 and 761). The oligomerization was
already observed but the masses obtained did not corre-
spond to identified oligomers in the literature [16].
An optimal temperature of 85 °C was thus chosen for
the rest of the study, associated with the catalyst B which
gave the best results.
The effect of the stirring speed on the oxidative cleavage at
OA/H2O2/[C5H5N(n-C16H33)]3{PO4[WO(O2)2]4}
(1/5/
0.02, by mol) was studied by varying the number of rev-
olutions per minutes (rpm) of the mechanical agitation
(Fig. 1).
The optimal speed agitation appeared to be around
400 rpm. Above this speed, a decrease in the acid yields
was observed. This could be explained by the introduction
of air bubbles in the media, which tends to decrease the
contact of the substrate/oxidant. The observation of the
increase of reaction volume with the agitation illustrated
this hypothesis. An agitation speed of 400 rpm was kept for
the rest of the study, as a stable emulsion, favoring the
active species transfer at the interface during the reaction,
Effects of the Catalyst Load
was obtained along with
afterwards.
a quite fast decantation
To pursue the investigation of the process, we then studied
the effect of the amount of catalyst on the yields. Experi-
ments using a catalyst concentration over a range of
1–4 mol% were carried out at 85 °C with an OA/H2O2
molar ratio of 1/5.
Kinetic Investigation
A kinetic investigation was then performed to complete this
study, using the optimized conditions. Samples of the reac-
tion media were taken at various times and analyzed by GL
C. The evolution in time of desired products (AA, PA) and
intermediate products (epoxide, diol, aldehydes) was studied
(Fig. 2). The sample, taken 10 min after the end of the H2O2
addition, showed that OA was almost totally consumed,
confirming that the peroxo anion is efficiently transferred.
Intermediates (epoxide, diol, aldehydes) were rapidly
formed but disappeared slightly slower with time, which
could explain the little shoulder on the AA and PA formation
curves (between 20 and 60 min). Noteworthy, the maximum
amounts of diol and aldehydes were obtained after the
Table 3 shows that a catalyst concentration of 2 mol%
gave the best results. Noteworthy, at higher concentrations,
Table 3 Effects of the catalyst load on the oxidative cleavage of OA
with the catalyst B
Catalyst load
(mol%)
vOA
(%)
YAA
(%)
YPA
(%)
Aldehyde
(%)
Epoxide
(%)
1
2
4
100.0 69.3
100.0 81.5
100.0 70.9
68.3
86.1
62.3
1.6
1.6
0.0
0.8
1.6
1.1
Reaction conditions: OA (65.0 mmol), H2O2 (325.0 mmol, 30 %),
85 °C, 5 h, 400 rpm
123

138
J Am Oil Chem Soc (2013) 90:133–140
Table 4 Effects of catalyst preparation and temperature treatment on
the oxidative cleavage of OA
Catalyst Catalyst preparation
method
Treatment
temperature
YAA
(%)
YPA
(%)
A
A
B
B
C
C
b
b
b
b
b
b
RT
72.9
77.6
71.8
81.5
71.3
75.7
73.9
80.9
70.9
86.1
77.2
80.7
Ice bath
RT
Ice bath
RT
Ice bath
Reaction conditions: OA (65.0 mmol), H2O2 (325.0 mmol, 30 %),
catalyst (1.2 mmol), 85 °C, 5 h, 400 rpm
Fig. 2 Relative areas evolution with time using OA (65.0 mmol),
H2O2 (325.0 mmol, 30 %), catalyst B (1.2 mmol), 85 °C, 400 rpm.
35 % mol of B. In order to have an efficient recycling for
the catalyst B, both the solid catalyst and the entire aqueous
phase needed to be reused to obtained good yields (74.1 %
of AA and 68.9 % of PA). The catalyst B has probably
evolved but it is still active under its new form in the
aqueous phase. This result is very interesting in terms of
sustainable chemistry (limitation of effluent wastes) and for
cost reduction (recovery).
AAA/Atotal (filled rectangles), APA/Atotal (filled triangles), Aepoxide/Atotal
(open diamonds),
Adiol/Atotal (filled circles), Anonanal/Atotal (2),
A9-oxo-nonanoic acid/Atotal (?). The end of the H2O2 addition was
considered as time 0
maximum amount of epoxide, which thus appears to be the
first intermediate of the intrinsic oxidation reaction.
In the literature, different pathways have been described
for the oxidative scission of oleic acid or of the methyl
ester (comprising one to four steps) [2, 11]. In our case, this
study clarified the mechanism: OA was oxidized to the
epoxide form, followed by the ring opening and the oxi-
dative cleavage. Thus, the species presented in the organic
phase were OA, epoxide, diol, aldehydes, AA and PA.
Generalization of the Oxidative Scission to Other
Substrates
Finally, the oxidative scission conditions were applied to
four other substrates (Table 5; Fig. 3).
The first substrate, 9,10-epoxyoctadecanoic acid, is an
epoxide. AA and PA were obtained in excellent yields
(respectively 87.3 and 86.5 %) using our optimized con-
ditions. To our knowledge, this is the first example of an
epoxide cleavage using this oxidative system. These results
Influence of the Treatment
In the previous papers reporting the use of catalysts A, B
and C, the extraction solvents used appeared to be harmful
(dichloromethane [10, 17] or diethyl ether [1, 9, 18]). In
contrast, ethyl acetate, which has already been used with an
other oxidant system [19], is safer and was thus chosen for
the optimization of the treatment after the oxidative
cleavage reaction. Two protocols with ethyl acetate were
tested: one at room temperature and the other one at a
temperature below 4 °C by cooling the reaction mixture in
an ice bath (Table 4).
Table 5 Oxidative cleavage of fatty substrates
Catalyst Substrate Conversion
(%)
Mono-acid
(%)
Di-acid
(%)
Ba
Bb
Bd
Be
Ba
LA
100.0
100.0
98.9
47.7
60.7
87.3
60.8
62.3
41.0c
49.8c
86.5
83.2
68.6
LA
Epoxide
RA
The extractions performed at low temperature gave
better results, with an increase in the yields of 5–7 %.
Indeed, it is well-known that the acids formed are partially
soluble in water, especially the di-acid (AA) (2.4 g L-1 at
20 °C) [20]. Cooling the crude reaction mixture allowed
their solubility in water to decrease and finally to get a better
extraction of the organic molecules from the aqueous phase.
In addition, solubility tests of freshly prepared catalysts
A and B were carried out and proved that at low temper-
atures (below 4 °C); catalysts A and B were almost
insoluble either in water or in ethyl acetate. After a low
temperature treatment, a filtration yields a recovery of
99.9
EO
99.9
Reaction conditions
a
Substrate (60 mmol), H2O2 (350 mmol, 5.8 equiv., 30 %), catalyst
(1.2 mmol, 2 mol%)
b
LA (22 mmol), H2O2 (220 mmol, 10.0 equiv., 30 %), catalyst
(0.9 mmol, 4 mol%)
c
The di-acid is exclusively AA as malonic acid is soluble in water
(solubility at 20 °C: 73 g L-1
)
d
Epoxide (51 mmol), H2O2 (255 mmol, 5 equiv., 30 %), catalyst
(1.2 mmol, 2 mol%)
e
RA (14 mmol), H2O2 (163 mmol, 30 %), catalyst (2 mol%)
123

J Am Oil Chem Soc (2013) 90:133–140
139
Fig. 3 Extension of the
oxidative cleavage to other
substrates
showed that the epoxide derivative is one of the interme-
diates of the oxidative cleavage of double bonds.
Our second choice was linoleic acid (cis-9, cis-12 oc-
tadecadienoic acid, LA). Indeed, this acid bears two double
bonds which can be cleaved leading to two diacids, azelaic
and malonic acid and a monoacid, hexanoic acid. Good
yields were obtained, thanks to the use of a double amount
of hydrogen peroxide and catalyst, to cleave both of the
double bonds of LA.
peroxo–tungsten complex Q3{PO4[WO(O2)2]4}, in the
presence of hydrogen peroxide as an oxidant and without
any organic solvent. The emulsifying system promoted
the OA reactivity in a biphasic medium through the
transfer of the active peroxo–tungsten complex. The sur-
factant power of fatty acids contributed to the stability of
the emulsion. The study of the mechanism showed the
successive formation of intermediate products (epoxides,
diol, aldehydes).
The following substrate was ricinoleic acid ((9Z,12R)-
12-hydroxyoctadec-9-enoic acid, RA). This fatty acid,
obtained from the castor oil, contains a hydroxyl on the
hydrocarbon chain. It is interesting to note that the
hydroxyl was mainly conserved: 3-hydroxynonanoic acid
and azelaic acid were obtained in good yields (respectively
83.2 and 60.8 %) [21].
It was proved that the catalytic activity of Q3{PO4
[WO(O2)2]4} depends on the type and the lipophily of the
associated quaternary ammonium salt Q?, X-. The study
of several parameters allowed us to provide optimized
conditions for the oxidative scission. Thus, high AA and
PA yields were obtained at 85 °C for 5 h with an OA/
H2O2/[C5H5N(n-C16H33)]3{PO4[WO(O2)2]4} molar ratio
of 1/5/0.02 associated with an in-situ protocol for the
preparation of the catalyst. This in-situ protocol allows us
to limit the loss of matter and the catalyst can be efficiently
recovered thanks to a treatment at low temperature. Finally,
this new process was generalized for the use of other
starting acids, demonstrating that these conditions can be
used to cleave several double bonds and even epoxides.
This clean process could be easily extended to other fatty
acids derivatives in order to obtain new short carbon chains
for bio-based products.
Thelastsubstratewasnotanacidbutafattyacidester:ethyl
oleate (EO). The cleavage was successfully achieved with
62.3 % of mono-methyl azelate and 68.9 % of PA The pro-
cess could be extended to other fatty acid esters derivatives.
Conclusion
This study revealed that the oxidative cleavage of OA
was successfully achieved using a catalytic amount of
123

140
J Am Oil Chem Soc (2013) 90:133–140
Acknowledgments This work was carried out as part of the project
12. Mouloungui Z., Melching E. (2004). Process for preparation of
fatty acids by in situ hydrolysis of the lipids contained in seeds of
a plant. FR Patent 2,843,970
´
Oleovision which is a program FUI (FUI-AAP7-OLEOVISION)
financed by the Region Midi-Pyrenees in France thanks to the Conseil
´
´ ´
regional, the FEDER and the Minister of Industry.
13. Salimon J, Salih N (2009) Oleic acid diesters: synthesis, char-
acterization and low temperature properties. Eur J Sci Res
32(2):216–222
14. Starks CM, Liotta CL, Halpem M (1994) Phase-transfer catalysis:
fundamentals, applications and industrial perspectives, phase-
transfer catalysts, 1st edn. Chapman and Hall, London
15. Timofeeva MN, Pai ZP, Tolstikov AG, Kustova GN, Selivanova
NV, Berdnikova PV, Brylyakov KP, Shangina AB, Utkin VA
(2003) Epoxidation of cycloolefins with hydrogen peroxide in the
presence of heteropoly acids combined with phase transfer cat-
alyst. Russ Chem Bull 52(2):480–486
16. Kozhevnikov IV, Mulder GP, Steverink-de Zoete MC, Oostwal
MG (1998) Epoxidation of oleic acid catalyzed by peroxo
phosphotungstate in a two-phase system. J Mol Cat A: Chem
134(1–3):223–228
17. Venturello C, D’Aloisio R (1988) Quaternary ammonium tetra-
kis(diperoxotungsto)phosphates(3-) as a new class of catalysts
for efficient alkene epoxidation with hydrogen peroxide. J Org
Chem 53(7):1553–1557
18. Antonelli E, D’Aloisio R, Gambaro M, Fiorani T, Venturello C
(1998) Efficient oxidative cleavage of olefins to carboxylic acids
with hydrogen peroxide catalyzed by methyltrioctylammonium
tetrakis(oxodiperoxotungsto)phosphate(3-) under two-phase
conditions. synthetic aspects and investigation of the reaction
course. J Org Chem 63(21):7190–7206
References
1. Turnwald SE, Lorier MA, Wright LJ, Mucalo MR (1998) Oleic
acid oxidation using hydrogen peroxide in conjunction with
transition metal catalysis. J Mater Sci Lett 17(15):1305–1307
¨
2. Kockritz A, Blumenstein M, Martin A (2010) Catalytic cleavage
of methyl oleate or oleic acid. Eur J Lipid Sci Technol
112(1):58–63
3. Telos SRL, Bader S, Masiello S (1999) Cosmetic composition
containing azelaic acid. WO Patent 9,917,714
4. Sefton J (2001) Method and composition for treating acne. US
Patent 6,262,117
5. Noureddini H, Rempe ML (1996) Pelargonic acid in enhanced oil
recovery. J Am Oil Chem Soc 73(7):937–941
¨
6. Kockritz A, Martin A (2008) Oxidation of unsaturated fatty acid
J Lipid Sci Technol
derivatives and vegetable oils. Eur
110(9):812–824
7. Zaldman B, Kisilev A, Sasson Y, Garti N (1998) Double bond
oxidation of unsaturated fatty acids. J Am Oil Chem Soc
65(4):611–615
8. Sato K (1998) Oxidation of sulfides to sulfoxides and sulfones
with 30 % hydrogen peroxide under organic solvent and halogen-
free conditions. Tetrahedron 57(13):2469–2476
9. Pai ZP, Khlebnikova TB, Mattsat YV, Parmon VN (2009) Cat-
alytic oxidation of fatty acids. I. Epoxidation of unsaturated fatty
acids. React Kinet Catal Lett 98(1):1–8
10. Pai ZP, Tolstikov AG, Berdnikova PV, Kustova GN, Khlebnik-
ova TB, Selivanova NV, Shangina AB, Kostrovskii VG (2005)
Catalytic oxidation of olefins and alcohols with hydrogen per-
oxide in a two-phase system giving mono and dicarboxylic acids.
Russ Chem Bull 54(8):1847–1854
19. Zimmermann F, Meux E, Mieloszynski J-L, Lecuire J-M, Ogeta
N (2005) Ruthenium catalysed oxidation without CCl4 of oleic
acid, other monoenic fatty acids and alkene. Tetrahedron Lett
46:3201–3203
20. Johnson RW (1989) Fatty acids in industry: processes, properties,
derivates, applications. Dibasic fatty acids. Marcel Dekker, New York
21. As Pai et al. mentioned, the secondary alcohol functionality is
less reactive than the primary alcohol functionality [9, 10]
¨
11. Kockritz A, Martin A (2011) Synthesis of azelaic acid from
J Lipid Sci Technol
vegetable oil-based feedstocks. Eur
113(1):83–91
123